This technology enables autonomous vehicles to plan high-level tasks and low-level motions in a unified, adaptive framework. By continuously sharing safety and efficiency information between planning layers, vehicles can make smarter decisions in complex, unpredictable urban environments. The result is safer navigation without sacrificing operational efficiency.
Background:
Autonomous vehicles must operate safely and efficiently in dynamic urban environments where traffic behavior is uncertain and constantly changing. Traditional systems separate high-level task planning from low-level motion planning, causing task plans to overlook real-time physical risks and motion plans to react conservatively without regard for broader goals. This disconnect results in inefficiencies, stalled maneuvers, or unsafe behaviors, and existing integrated approaches from robotics struggle to adapt to the multi-agent uncertainty inherent in urban driving.
Technology Overview:
The invention integrates symbolic task planning with continuous motion planning through a closed feedback loop. A task planner generates action sequences that are translated into navigation goals by a plan manager, while a motion planner produces collision-free trajectories. A safety estimator assesses the safety of each planned action based on available control signals and predicted traffic behavior. Safety feedback flows back to the task planner, enabling dynamic replanning that balances efficiency with real-time risk.
Advantages:
• Improves decision making by linking task choices directly to real-time motion safety • Reduces unsafe maneuvers through continuous safety assessment • Enhances efficiency by avoiding overly conservative behavior • Adapts planned actions dynamically to changing traffic conditions
Applications:
• Autonomous urban ride-hailing • Autonomous delivery vehicles • Advanced driver assistance systems • Fleet coordination for urban mobility services
Intellectual Property Summary:
• United States – 12,384,410 – Issued 8/12/2025 • United States – 19/295,936 – Utility – Filed 8/11/2025
Stage of Development:
Prototype
Licensing Status:
This technology is available for licensing.
Licensing Potential:
Strong potential for autonomous vehicle developers, mobility service providers, and automotive technology companies seeking safer, more efficient planning systems for complex urban driving environments.
Additional Information:
Information available upon request.
Inventors: Yan Ding, Shiqi Zhang
Alternate NCS Title: Integrated Planning for Safer, Smarter Autonomous Urban Driving